I would agree that a Ping < 1ms is not really an accurate representation. Ping is not really a good measure of service. Ping has a number of dependencies including the number of steps to the pinged device. How links handle ping. The packet size of the ping. For example pinging from a Cisco Router can deliver a different result than ping from a PC which can be different than pinging from a Unix based system. Why, I here you ask, each has different packet size.

The best way to measure performance is by performing a package download from a known server on a known path. I see a number of people here use OOKLA this is most likely a pretty fair measurement but it is probably best to ensure it is measuring via the correct server (usually this is ok but not always).

That was my experience and teachings but I am getting old and have been retired for over 7 years.

Incidentally: using Ookla to one server I get a ping of 12ms download 53Mbs Up 18Mbs (not bad for a 50Mbs link). Go to another server a ping of 13ms and speeds between 51 - 52Mbs Down but Up however is 8.8Mbs. These figures pretty much explain what I’m getting at.

My two bobs worth

Tony


Last edited by Teunis; 11/19/20 08:31 PM.
HP i7-4770 16GB 1TB SSD, Win 10 Home,
Focusrite 2i2 3rd Gen, Launchkey 61, Maton CW80, Telecaster, Ovation Elite TX, Yamaha Pacifica 612
BB 2022(912) RB 2022(2), CakeWalk, Reaper 6, Audacity, Melodyne 5 Editor, Izotope Music Production Suite 4.1